Data transmission method and system based on edge computing
By constructing a channel-motion coupling fluctuation index and a delay sensitivity weighted backlog coefficient, and dynamically adjusting the control parameters of the Lyapunov optimization algorithm, the problem of data queue backlog in the communication command vehicle was solved, and real-time data transmission in complex environments was realized.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- GUANGZHOU WEIBANG VEHICLE EQUIP
- Filing Date
- 2026-02-26
- Publication Date
- 2026-05-26
Smart Images
Figure CN122093484A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data transmission technology. More specifically, this invention relates to a data transmission method and system based on edge computing. Background Technology
[0002] Edge computing-based data transmission refers to the technology in which, within the network architecture of a communication command vehicle, the massive amounts of heterogeneous sensing data (such as high-definition surveillance video, radar signals, and sensor readings) generated by the vehicle-mounted terminals are wirelessly routed and transmitted to edge computing servers deployed locally or in nearby nodes (such as accompanying drones or portable base stations) for processing. When communication command vehicles are performing emergency rescue or field operations, due to the complex on-site environment, limited public network bandwidth, or unstable links, directly transmitting all raw data to the cloud for processing would result in unacceptable latency. Edge computing-based data transmission enables localized offloading and real-time processing of data at the vehicle's edge, significantly reducing the transmission pressure on the core network. This plays a crucial role in improving the situational awareness speed of the command system and ensuring the real-time nature of command issuance.
[0003] Currently, Lyapunov optimization algorithms are widely used to address the data transmission resource allocation and load balancing problems in edge computing scenarios. Based on the drift-penalty theory framework, this algorithm transforms long-term time-averaged constraints (such as average energy consumption and average latency) into deterministic optimization subproblems for each time slot. It boasts significant advantages, including no need for prior knowledge of future channel state information (CSI), low computational complexity, and theoretically guaranteed queue stability. However, communication command vehicles face highly dynamic wireless communication environments during movement. Channel quality fluctuates drastically with high vehicle speed and terrain obstruction, and data traffic triggered by emergency tasks exhibits strong bursts. Existing Lyapunov optimization algorithms typically use a fixed control parameter to balance energy consumption and queue backlog (i.e., latency). This static parameter configuration lacks dynamic adaptability to environmental changes. In the event of a sudden deterioration in channel conditions or an explosive increase in data, fixed control parameters cannot quickly adjust the priority of the transmission strategy. This leads to excessive backlog of data queues in the process of minimizing energy consumption, resulting in technical problems such as excessively high data transmission latency. This fails to meet the stringent requirements of timely real-time communication for communication command vehicles in emergency situations. Summary of the Invention
[0004] This invention provides a data transmission method and system based on edge computing, aiming to solve the problem in related technologies where fixed control parameters cannot quickly adjust the priority of the transmission strategy, resulting in excessive data queue backlog in the process of minimizing energy consumption, thus causing excessive data transmission latency.
[0005] In a first aspect, the present invention provides a data transmission method based on edge computing, comprising: acquiring vehicle motion data and channel state information of a communication channel, wherein the motion data includes acceleration and the channel state information includes signal-to-noise ratio (SNR); calculating a channel-motion coupling fluctuation index based on multiple SNRs and accelerations within a set sliding window length; wherein the channel-motion coupling fluctuation index is positively correlated with the sum of the absolute differences between adjacent SNR values within the sliding window and with the root mean square value of acceleration within the sliding window; acquiring the current physical queue length of an edge node buffer and calculating a delay sensitivity weighted accumulation coefficient; wherein the delay sensitivity weighted accumulation coefficient is positively correlated with the physical queue length, the channel-motion coupling fluctuation index, and the data packet arrival rate within the window; setting adaptive control parameters based on the delay sensitivity weighted accumulation coefficient and adjusting a transmission control algorithm using the adaptive control parameters, wherein the adaptive control parameters are inversely proportional to the delay sensitivity weighted accumulation coefficient; and performing data transmission according to the adjusted transmission control algorithm.
[0006] Furthermore, the specific method for calculating the channel-motion coupling fluctuation index is as follows: In the formula, Indicates time Channel-motion coupling fluctuation index; This indicates the length of the sliding window; Indicates time Signal-to-noise ratio; Indicates time The acceleration; The acceleration due to gravity is a constant. By nonlinearly coupling the change in signal-to-noise ratio (SNR) with vehicle acceleration, the perturbation effect of vehicle mechanical motion on wireless channel quality can be accurately quantified. This calculation method uses acceleration as a gain coefficient to amplify the weight of SNR fluctuations during severe vehicle jolting or maneuvering. This allows for a more sensitive detection of precursors to adverse communication environments than simply monitoring channel conditions, providing a highly sensitive input basis for subsequent parameter adjustments.
[0007] Furthermore, the specific method for calculating the time delay sensitivity weighted product coefficient is as follows: In the formula, Indicates time The time delay sensitivity weighted product coefficient; Indicates time The physical queue length of the buffer; Represented by natural constant An exponential function with base 0; For a moment Channel-motion coupling fluctuation index; For a moment Average packet arrival rate within the time window; A preset reference throughput threshold is set for the system. By introducing an exponential function to construct a latency-sensitive weighted backlog coefficient, the limitation of traditional algorithms that only linearly depend on the physical queue length is overcome. In harsh environments with large channel fluctuations (high coupling index), this index utilizes the amplification effect of exponential characteristics to quickly calculate extremely high virtual backlog values even when the physical queue is short. This provides early warning of congestion risks, forcing the system to adopt aggressive transmission strategies before actual network paralysis, effectively avoiding data cliff-like accumulation in highly dynamic scenarios.
[0008] Furthermore, the calculation method for the adaptive control parameters is as follows: In the formula, Indicates time Adaptive drift penalty control parameters used to optimize the algorithm; and These are the theoretical optimal values for time delay and energy saving, respectively; For a moment The time delay sensitivity weighted product coefficient; To adjust the sensitivity coefficient, the response speed of the control parameters to changes in the backlog index was controlled. An inverse proportional function relationship between the control parameters and the backlog index was established, enabling smooth and rapid switching between energy-saving and low-latency modes. When the backlog risk increases, the control parameters approach the theoretical optimal value for latency, forcing the system to clear the queue at the cost of energy consumption; when the environment is stable, the parameters rise to save energy. This dynamic mapping mechanism solves the problem of sluggish response of static parameters under sudden conditions, achieving millisecond-level strategy adjustment.
[0009] Furthermore, acquiring vehicle motion data and communication channel state information includes: using an inertial measurement unit (IMU) to collect the vehicle's instantaneous velocity and vertical acceleration components; calibrating the instantaneous velocity using a GPS positioning module; and monitoring the channel state information, including the signal-to-noise ratio (SNR) sequence and bit error rate, using the physical layer interface of the communication module. By integrating the IMU, GPS positioning module, and communication physical layer interface, multi-dimensional synchronous perception of the physical world's motion state and the electromagnetic space channel state is achieved. Utilizing GPS to calibrate IMU data and combining it with the physical layer SNR ensures the accuracy and comprehensiveness of the input data, providing a reliable data foundation for the algorithm to accurately determine the current operating condition.
[0010] Furthermore, the process includes preprocessing the motion data and channel state information. This preprocessing includes smoothing the acceleration using a Kalman filter algorithm and normalizing the signal-to-noise ratio (SNR) to map its value range to the [0, 1] interval. Kalman filtering eliminates high-frequency noise generated by vehicle vibration, and SNR normalization eliminates differences between data of different dimensions. This avoids frequent oscillations in control parameters due to instantaneous sensor errors or minor disturbances, ensuring the numerical stability and robustness of the adaptive control algorithm in practical applications, and enabling the system to respond only to real trend changes.
[0011] Furthermore, the transmission control algorithm is a Lyapunov optimization algorithm, and the adaptive control parameters are used to adjust the weights of the power-related penalty terms in the Lyapunov optimization algorithm. Specifically, the adaptive parameters are applied to adjust the weights of the power (energy consumption)-related penalty terms in the Lyapunov optimization algorithm. This retains the theoretical advantage of the Lyapunov algorithm—that it does not require prior knowledge of future channel information and can guarantee queue stability—while also giving it the flexibility to dynamically adjust priorities, achieving an optimal balance between theoretical stability and practical anti-interference capability.
[0012] Furthermore, the reference throughput threshold is set at 40% to 60% of the system's maximum available bandwidth.
[0013] Furthermore, the sensitivity coefficient ranges from 1.2 to 1.8.
[0014] In a second aspect, an edge computing-based data transmission system is also provided, comprising a processor and a memory, characterized in that the memory stores a computer program, and the processor executes the computer program to implement the edge computing-based data transmission method described in any of the preceding claims.
[0015] Beneficial Effects: By fusing vehicle acceleration and channel signal-to-noise ratio to construct a channel-motion coupling fluctuation index, and using this index to calculate a latency-sensitive weighted backlog coefficient reflecting the urgency of real-world services, the adaptive parameters of the transmission control algorithm are dynamically adjusted. This enables the system to perceive the potential channel degradation risk caused by drastic vehicle movement, automatically reducing its focus on energy conservation and prioritizing real-time data transmission when the environment deteriorates or data bursts occur. This significantly reduces transmission backlog and latency in highly dynamic edge computing scenarios. Attached Figure Description
[0016] Figure 1 This is a schematic flowchart illustrating a data transmission method according to an embodiment of the present invention;
[0017] Figure 2 This illustration shows a comparison of data backlog between the traditional algorithm and the improved algorithm when transmitting data according to the present invention. Figure 3 This illustration shows a comparison of latency when transmitting data using the conventional algorithm and the improved algorithm according to the present invention. Detailed Implementation
[0018] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.
[0019] like Figures 1 to 3 As shown, S101: Acquisition and preprocessing of multidimensional data.
[0020] In this embodiment, for the high-dynamic scenario of the communication command vehicle, multi-dimensional data is synchronously collected through sensors and communication modules deployed on the vehicle body. Specifically, the inertial measurement unit (IMU) is used to collect vehicle motion data at various times, including instantaneous velocity and vertical acceleration components; the GPS positioning module is used to assist in calibrating the vehicle's position and speed information; the physical layer detection interface of the communication module is used to monitor the channel state information at various times in real time, including the channel signal-to-noise ratio sequence and bit error rate; simultaneously, the queue length of the current data buffer, the data packet arrival rate, and the data packet lifetime are collected at various times through the data link layer interface of the edge gateway.
[0021] After data acquisition, the raw data is preprocessed to eliminate noise interference. First, a Kalman filter algorithm is used to smooth the instantaneous speed and acceleration data of the vehicle to remove high-frequency noise caused by vehicle mechanical vibration. Second, the signal-to-noise ratio data is normalized and mapped to... Intervals were defined to eliminate dimensional differences. Finally, box plot analysis was employed or... The criteria perform outlier detection on the data queue length, removing extreme points caused by transient system failures. The processed data is then aligned by timestamps and constructed into a time series vector to provide input for subsequent indicator calculations.
[0022] S102: Construct the channel-motion coupling fluctuation index.
[0023] In communication command vehicle scenarios, vehicles often need to maneuver at high speeds across rugged roads or complex terrain such as urban ruins. This intense motion causes a sharp increase in multipath effects and Doppler shift in wireless signals, resulting in nonlinear and drastic fluctuations in channel quality. Existing data transmission algorithms often only focus on the average signal-to-noise ratio (SNR), neglecting the strong coupling between the degree of SNR change and the vehicle's motion state. When the vehicle is under high acceleration or experiencing severe bumps, the channel quality is highly likely to plummet in the next instant. This motion-induced channel uncertainty leads to a surge in data retransmission rates, potentially congesting the transmission queue. Therefore, it is necessary to construct an index that integrates vehicle motion intensity and channel change rate to quantify this unique and harsh environmental characteristic. The more pronounced this characteristic, the more unreliable the current communication environment, and the more redundant resources the algorithm needs to reserve.
[0024] Before constructing this metric, the preprocessed data needs to be divided into sliding window segments, with the sliding window size set to [value missing]. Extraction deadline The signal-to-noise ratio (SNR) sequence within the sliding window and the corresponding synthetic vehicle acceleration sequence are calculated. The absolute value of the first-order difference of the SNR within the sliding window is calculated to reflect the instantaneous rate of change of the channel. Simultaneously, the root mean square (RMS) value of the acceleration within the sliding window is calculated to reflect the intensity of the vehicle's motion. Based on the above analysis, a time-sequence... Channel-motion coupling fluctuation index The formula is as follows: In the formula, Indicates time The channel-motion coupling fluctuation index, the larger the value, the more severe the environment; Indicates the time step of the sliding window; Indicates time Normalized signal-to-noise ratio value Relative to the current time Push forward One time unit; Indicates time The vehicle's composite acceleration value; For reference, the gravitational acceleration constant is typically taken as 9.8. It is used to perform dimensionless correction on the acceleration term.
[0025] As can be seen from the formula, This reflects the amplitude of changes in the channel itself, while the acceleration term within parentheses, as a gain coefficient, reflects the amplifying effect of the vehicle's violent movement on channel uncertainty. When the communication command vehicle moves at high speed in adverse road conditions, the acceleration... The signal-to-noise ratio (SNR) increases, and the multipath effect causes a sharp jump in the signal-to-noise ratio. These two factors are multiplied and added together, resulting in a change in the calculated time. Channel-motion coupling fluctuation index The increase is significant, thus accurately reflecting the high-risk characteristics of the current communication environment.
[0026] S103: Construct the delay sensitivity weighted backlog coefficient.
[0027] In a communication command vehicle scenario, if the external environment is extremely unstable, i.e., the channel-motion coupling fluctuation index is large, the risk of data packets waiting to be transmitted in the data queue will increase rapidly. This is because once the channel deteriorates, retransmission will consume a significant amount of time, preventing subsequent urgent intelligence from being sent. Therefore, the physical queue length alone is insufficient to reflect the true urgency of the business. A unique characteristic is that the same queue length, acceptable in a stable environment, becomes a fatal congestion hazard in a drastically fluctuating environment. We need to construct an indicator that uses environmental volatility to weight the physical queue length, thereby obtaining an effective backlog level that reflects the actual transmission pressure.
[0028] Before constructing this metric, time points need to be obtained. Real-time physical queue length and time of edge node data buffer The average arrival rate of data packets. The physical queue length is compared with the channel-motion coupling fluctuation index obtained in step S102. Perform time-series alignment to ensure that fluctuation metrics under the current environmental conditions are used. Based on this, construct the time series. Delay sensitivity weighted product coefficient The formula is as follows: ;In the formula, Indicates time The time delay sensitivity weighted product coefficient; Indicates time The physical queue length of the buffer; Represented by natural constant An exponential function with base 0; For a moment Channel-motion coupling fluctuation index; For a moment Average packet arrival rate within the time window; The preset reference throughput threshold for the system can be set to half of the network card's maximum bandwidth, or to 40% to 60% of the system's maximum available bandwidth.
[0029] This formula utilizes the amplification property of the exponential function for risk warning. Logical derivation shows that as the channel-motion coupling fluctuation exponent increases, the value of the exponential term rises rapidly. This means that in adverse environments, even without a significant increase in physical queue length, the calculated delay sensitivity weighted accumulation coefficient will be significantly amplified. This reflects that the transmission difficulty and potential delay cost per unit of data are far higher in unstable channels than in stable channels, enabling the algorithm to detect potential congestion crises caused by environmental degradation and thus trigger adjustments to the transmission strategy in advance.
[0030] S104: Construct adaptive drift penalty control parameters.
[0031] Based on the above time Delay sensitivity weighted product coefficient This needs to be mapped to the core control parameters in the Lyapunov optimization algorithm. In the traditional drift-penalty algorithm, the parameters balance energy consumption and queue stability. Larger parameter values tend to conserve energy, but this increases queue backlog; smaller parameter values tend to clear the queue quickly, but this results in higher energy consumption. In the unique scenario of the communication command vehicle, a high latency-sensitive weighted backlog coefficient indicates a harsh environment and a high risk of backlog. The algorithm must be forced into a low-latency mode, i.e., rapidly reducing the control parameter value, sacrificing energy consumption to ensure the transmission of critical intelligence; conversely, increasing the control parameter value conserves energy. Therefore, an index, i.e., a dynamic parameter, needs to be constructed that has a non-linear inverse relationship with backlog risk, enabling rapid strategy switching.
[0032] Before constructing this metric, it is necessary to define the allowable range of parameter variations for the algorithm. ,in This is the theoretical optimal value considering only energy saving. This is the theoretical optimal value when only time delay is considered.
[0033] Based on this, construct the time. Adaptive drift penalty control parameters The formula is as follows: ;In the formula, Indicates time Adaptive drift penalty control parameters used to optimize the algorithm; and These are the theoretical optimal values for delay and energy saving, respectively; For a moment The time delay sensitivity weighted product coefficient; The sensitivity coefficient is used to control the response speed of the parameter as the backlog index changes. It is a constant value, for example, the value range is 1.2~1.8, and in this embodiment, the value is 1.5.
[0034] This formula employs an inverse proportional fractional structure, establishing a nonlinear inverse proportional relationship between backlog risk and control parameters. At time... Delay sensitivity weighted product coefficient When the environment is harsh or there is severe backlog in the queue, the denominator increases rapidly, causing the fractional terms to approach zero. Approaching This forces the Lyapunov optimization algorithm to significantly reduce its focus on energy consumption and prioritize queue transmission, thus entering a low-latency mode. Conversely, when the environment is stable and the queue is relatively empty, Smaller rebounded to In the vicinity, the algorithm tends to be energy-efficient.
[0035] S105: Algorithm Improvement and Data Transmission Execution.
[0036] The time calculated based on step S104 Adaptive drift penalty control parameters An improvement is made to the original Lyapunov optimization algorithm. In each transmission time slot... Construct an optimization objective function that includes drift and penalty, i.e., minimize the following objective: In the formula, For transmission power; For the transmission rate related term, in the traditional Lyapunov drift plus penalty model, this is usually associated with the queue drift term, that is, it is related to the backlog and the difference between the input and output rates; Indicates time The physical queue length of the buffer. Unlike traditional algorithms, the control parameter here is no longer a constant, but rather a time value calculated from the preceding steps. Adaptive drift penalty control parameters .
[0037] The system solves the above optimization problem based on the current channel state information to derive the optimal transmission power and bandwidth allocation scheme for the current time slot. Because... It already incorporates comprehensive considerations of vehicle motion status, channel fluctuations, and queuing risks, and the resulting transmission scheme can automatically and smoothly switch between aggressive transmission and conservative energy saving.
[0038] After obtaining the optimal transmission power and bandwidth allocation scheme, the vehicle-mounted edge gateway of the communication command vehicle invokes the underlying hardware driver to transmit signals according to the calculated power and maps the data packets to be transmitted onto the corresponding subcarriers for transmission to the edge server. Simultaneously, the system monitors transmission feedback in real time; if the transmission in the current time slot is successful, the corresponding data packet is removed from the queue, and the queue status for the next time slot is updated. This triggers a new round of data collection and parameter calculation.
[0039] The present invention also provides a data transmission system based on edge computing. The system includes a processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement the data transmission method based on edge computing according to the first aspect of the present invention.
[0040] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and therefore will not be described in detail here.
[0041] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions stored or otherwise maintained on such a computer-readable medium.
[0042] The embodiments described above are merely examples of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.
Claims
1. A data transmission method based on edge computing, characterized in that, include: Acquire vehicle motion data and communication channel state information, wherein the motion data includes acceleration and the channel state information includes signal-to-noise ratio; Based on multiple signal-to-noise ratios and accelerations within a set sliding window length, a channel-motion coupling fluctuation index is calculated; the channel-motion coupling fluctuation index is positively correlated with the sum of the absolute differences between adjacent signal-to-noise ratio values within the sliding window and with the root mean square value of acceleration within the sliding window; Obtain the current physical queue length of the edge node buffer and calculate the latency sensitivity weighted backlog coefficient; the latency sensitivity weighted backlog coefficient is positively correlated with the physical queue length, the channel-motion coupling fluctuation index, and the data packet arrival rate within the window; Adaptive control parameters are set based on the time-delay sensitivity weighted integration coefficient, and the transmission control algorithm is adjusted using the adaptive control parameters. The adaptive control parameters are inversely proportional to the time-delay sensitivity weighted integration coefficient. Data transmission is performed according to the adjusted transmission control algorithm.
2. The data transmission method based on edge computing according to claim 1, characterized in that, The specific method for calculating the channel-motion coupling fluctuation index is as follows: ； In the formula, Indicates time Channel-motion coupling fluctuation index; This indicates the length of the sliding window; Indicates time Signal-to-noise ratio; Indicates time The acceleration; is the gravitational acceleration constant.
3. The data transmission method based on edge computing according to claim 1, characterized in that, The specific method for calculating the time delay sensitivity weighted integration coefficient is as follows: ； In the formula, Indicates time The time delay sensitivity weighted product coefficient; Indicates time The physical queue length of the buffer; Represented by natural constant An exponential function with base 0; For a moment Channel-motion coupling fluctuation index; For a moment Average packet arrival rate within the time window; This is a preset reference throughput threshold for the system.
4. The data transmission method based on edge computing according to claim 1, characterized in that, The specific method for calculating the adaptive control parameters is as follows: ； In the formula, Indicates time Adaptive drift penalty control parameters used to optimize the algorithm; and These are the theoretical optimal values for time delay and energy saving, respectively; For a moment The time delay sensitivity weighted product coefficient; The sensitivity coefficient is used to control the response speed of parameters as the backlog index changes.
5. The data transmission method based on edge computing according to claim 1, characterized in that, Acquire vehicle motion data and communication channel state information, including: The instantaneous velocity and vertical acceleration components of the vehicle are acquired using an inertial measurement unit; the instantaneous velocity is calibrated using a GPS positioning module. The channel state information, including the signal-to-noise ratio sequence and bit error rate, is monitored using the physical layer interface of the communication module.
6. The data transmission method based on edge computing according to claim 1, characterized in that, It also includes preprocessing the motion data and channel state information, the preprocessing including: The acceleration is smoothed using a Kalman filter algorithm, and the signal-to-noise ratio is normalized to map its numerical range to the [0, 1] interval.
7. The data transmission method based on edge computing according to claim 1, characterized in that, The transmission control algorithm is a Lyapunov optimization algorithm, and the adaptive control parameters are used to adjust the weights of the power-related penalty terms in the Lyapunov optimization algorithm.
8. The data transmission method based on edge computing according to claim 3, characterized in that, The reference throughput threshold is set at 40% to 60% of the system's maximum available bandwidth.
9. The data transmission method based on edge computing according to claim 4, characterized in that, The sensitivity coefficient ranges from 1.2 to 1.
8.
10. A data transmission system based on edge computing, comprising a processor and a memory, characterized in that, The memory stores a computer program, and the processor executes the computer program to implement the edge computing-based data transmission method as described in any one of claims 1-9.